Download Organelle origins: Energy-producing symbionts

Dispatch
R315
Organelle origins: Energy-producing symbionts in early eukaryotes?
Mitchell L. Sogin
The discovery that Trichomonas vaginalis, an early
diverging protist that lacks mitochondria but has
energy-producing hydrogenosomes, makes bacteriallike heat shock proteins suggests that symbionts
ancestral to mitochondria and hydrogenosomes were
present at early stages of eukaryote evolution.
Address: Bay-Paul Center for Comparative Molecular Biology and
Evolution, Marine Biological Laboratory, Woods Hole, Massachusetts
02543, USA.
Electronic identifier: 0960-9822-007-R0315
are micro-aerophilic, and most have adopted parasitic
lifestyles; none contains a recognizable mitochondrion, but
trichomonads possess hydrogenosomes. This suggested
that, if the rRNA-based trees were accurate reflections of
evolutionary history, then inventories of the ancestral features of early eukaryotes should not include mitochondria.
Four recent studies of heat shock proteins [5–8] challenge
this view of mitochondrial origins. The hydrogenosomes of
Trichomonas vaginalis contain heat shock proteins similar to
those found in mitochondria; these organelles may be
descendants of the ‘first mitochondria’ [9].
Current Biology 1997, 7:R315–R317
© Current Biology Ltd ISSN 0960-9822
The transition from a prokaryotic to a eukaryotic cell architecture generated one of the largest discontinuities in our
evolutionary history. Unlike bacteria and archaebacteria,
eukaryotes retain their genetic information, replication and
transcription machinery within a nuclear envelope, and
have an endomembrane-mediated system of protein
sorting and transport. Most eukaryotes compartmentalize
aerobic energy metabolism within mitochondria, the only
exceptions being amitochondriate protists, which produce
energy by fermentation in the cytoplasm or within
organelles called hydrogenosomes (Fig. 1) [1]. The innovation of energy-producing organelles forever altered metabolism in the eukaryotic line of descent. To understand when
this occurred, it is necessary to know the phylogenetic distribution of protists that lack mitochondria, as well as the
placement of eukaryotes that contain hydrogenosomes.
The origins of mitochondria and plastids are inextricably
tied to symbiotic relationships between eukaryotes and
particular bacterial groups. Ultrastructural, biochemical and
molecular sequence data all show that mitochondria share a
common evolutionary history with members of the alphadivision of purple photosynthetic bacteria (a-proteobacteria), and that cyanobacteria were the progenitors of
Figure 1
Nearly thirty-five years ago, the introduction of electron
microscopy and new techniques for exploring ultrastructure radically transformed views of protist evolution.
Instead of three or four classes of protists, sixty or seventy
major lineages were recognized. The relative branching
patterns of these lineages, however, remained speculative
until the construction of molecular trees [2]. The ensuing
studies, largely based on sequences of ribosomal RNAs
(rRNAs), confirmed the ultrastructure classifications, but
also provided quantitative measures of genetic relatedness
that can be used to infer relative branching orders.
Three protist groups, the diplomonads, microsporidia and
trichomonads, are basal to all other eukaryotes [3]. Their
divergence is followed by a series of independent protist
branchings, and then by the nearly simultaneous divergence of animals, plants, fungi and several other protist
groups. These suddenly diverging taxa form the ‘crown’ of
the eukaryotic tree [4]. The earliest diverging eukaryotes
A hydrogenosome of Trichomonas foetus, a close relative of the protist
T. vaginalis discussed in the text, showing the characteristic marginal
plate (m) and dense core (c). (Photograph courtesy M. Müller and H.
Shio, reproduced with permission from [11].)
R316
Current Biology, Vol 7 No 5
plastids. (Support for the endosymbiotic origins of other
organelles, such as basal bodies, is more tenuous.) Here, I
shall discuss the origin of hydrogenosomes in micro-aerotolerant protists and what that might tell us about the
presence or absence of mitochondria in the earliestbranching eukaryotes.
Hydrogenosomes are factories for anaerobic energy production. They use the enzymes pyruvate:ferredoxin
oxidoreductase and hydrogenase — which are characteristic of anaerobic bacteria — and the substrates pyruvate or
malate to produce ATP, acetate, carbon dioxide and large
amounts of hydrogen [1]. These organelles are doublemembrane structures which divide by fission, lack DNA
and contain approximately 300 different proteins. Organisms that contain hydrogenosomes are unlike other
eukaryotes, as they are amitochondriate and generally live
as facultative anaerobes. Yet their numerous ultrastructural and biochemical similarities suggest that hydrogenosomes and mitochondria have a shared evolutionary
history. Alternatively, these common features might be
superficial: hydrogenosomes may have arisen from independent endosymbioses between anaerobic bacteria and
eukaryotes.
To differentiate between these various possibilities, several
research groups [5–8] have searched the Trichomonas
vaginalis genome for coding regions that encode proteins
related to those found in mitochondria. They looked in particular for genes encoding proteins related to the three
main types of mitochondrial heat shock protein (Hsp) —
Hsp10, Hsp60 and Hsp70 — because sequences of these
genes are particularly useful for inferring bacterial phylogeny. As trichomonads contain hydrogenosomes but not
mitochondria, the identification of Hsp genes encoding
hydrogenosome Hsps related to those of mitochondria
would constitute evidence that these energy-producing
organelles have a shared common ancestry.
Some Hsp genes are induced by environmental stress,
others encode molecular chaperones that stabilize partially folded proteins during their translation or intracellular transport. Hsp genes are present in bacterial and
nuclear genomes, but not in those of eukaryotic
organelles or (with the exception of Hsp70) archaebacteria; the Hsps that localize to eukaryotic organelles are
nuclear-encoded and transported to their appropriate destinations. Bacteria and archaebacteria have a single Hsp70
gene, but eukaryotes generally have multiple Hsp70
genes encoding distinct proteins that localize in the
endoplasmic reticulum, cytosol and mitochondria or
hydrogenosomes. The assumption is that nuclear genes
for organellar proteins migrated from the organellar
genome at some point during eukaryotic evolution after
the acquisition of the endosymbiont from which the
organelle is thought to be derived.
If mitochondria and hydrogenosomes have a common evolutionary history, the genes encoding their Hsps should be
phylogenetically related. The presence of mitochondrionlike Hsp genes in deeper branches of the eukaryotic tree
would provide presumptive evidence that the earliest
diverging eukaryotes had mitochondria. This argument
has been applied to the amitochondriate Entamoebae,
which is presumed to be on a lineage of protists that once
had mitochondrion-like organelles [10]. By taking advantage of conserved sequences in Hsp genes, putative homologues from T. vaginalis have now been cloned, sequenced
and subjected to phylogenetic analysis [5–8]. In all cases, a
battery of inference techniques place the trichomonad Hsp
sequences on the lineage that contains mitochondria. For
example, tree topologies that unite mitochondrial and trichomonad hydrogenosome Hsp60s are statistically robust;
the Hsp60 trees also show specific relationships between
mitochondria and rickettsial members of the a-proteobacteria. Similarly, the Hsp70 trees demonstrate a convincing
relationship between of the trichomonad hydrogenosomes
with mitochondria, but not with a-proteobacteria.
Mitochondria and hydrogenosomes must thus have a
common evolutionary history — but at what point did
their lineages diverge? Hydrogenosomes may have
evolved from an endosymbiont that was ancestral to
mitochondria [8], but this seems unlikely because
hydrogenosomes are present in anaerobic fungi and rumen
ciliates. In phylogenetic trees based on molecular data, the
anaerobic fungi and rumen ciliates are ‘crown group’ taxa
which diverged well after the endosymbiotic event that
produced eukaryotic mitochondria. Hydrogenosomes were
thus present in late-diverging, as well as early-diverging,
eukaryotes. This means that, if we interpret all
hydrogenosomes as being derived from an ancestral symbiont that also gave rise to mitochondria, the transition
from symbiont to hydrogenosomes or to mitochondria
would have occured many times. It seems more likely that
hydrogenosomes are, in fact, highly reduced and modified
descendants of mitochondria.
A second possibility is that two or more independent
endosymbiotic events occurred, one giving rise to the trichomonad hydrogenosomes and the other the mitochondria found in extant eukaryotes. The two endosymbionts
would have been different, but closely related, a-proteobacteria. This would explain the similarity between
mitochondrial and trichomonad Hsp genes. The trichomonad endosymbiont might have been transient and
merely allowed transfer of several bacterial genes, including those for heat shock proteins, to the nuclear genome.
Bacterial symbionts occur in many phylogenetically
diverse protists including amitochondriate pelobionts and
the diplomonad Giardia lamblia. Endosymbiosis clearly
did not occur just once during the evolution of eukaryotes; this process continues today, and sometimes even
Dispatch
involves two different species of eukaryote. Whenever
endosymbiotic relationships are established, a window of
opportunity is opened for transferring genetic information between genome compartments. This mechanism
could explain how Hsp genes entered the nuclear
genome of early diverging lineages such as the
trichomonads.
R317
medical and economical importance of these amitochondriate, parasitic protists. There may be earlier-diverging
eukaryotes, perhaps living in anaerobic or extreme
environments, awaiting discovery and placement in phylogenetic frameworks. Will they too have molecular evidence
of hydrogenosomes or mitochondria?
References
A third possibility is that hydrogenosomes are highly
reduced descendants of mitochondria. This implies mitochondria were present before the divergence of trichomonads. If these energy-producing organelles were
present in the earliest branches of the eukaryotic tree, we
would expect congruence between Hsp phylogenies and
other nuclear gene trees. The Hsp10 data base is too small
to address this issue, and the Hsp70 gene trees are not well
resolved. Furthermore, the Hsp70 trees do not demonstrate a specific relationship between the mitochondrial
lineage and a particular division of a-proteobacteria. The
Hsp60 gene phylogeny shows better resolution with a
sparse branching pattern that resembles the eukaryotic
rRNA phylogenies and also provides modest support for
the view that mitochondria and a-proteobacteria have a
common evolutionary history.
The sequencing of more Hsp genes may eventually
clarify the number of endosymbiotic events that led to
the formation of mitochondria and hydrogenosomes. With
the data already available, however, we can make reasonable estimates about relative timing for the origins of
energy producing organelles. The introduction of mitochondria and/or hydrogenosomes into the eukaryotic cell
must have occurred after the nuclear lineage had
diverged from the prokaryotic world. If this were not the
case, genes for mitochondrial proteins would be more
similar to ‘bona fide’ eukaryotic genes than to genes of
bacteria or archaebacteria (the latter are now considered
to be a sister group to all eukaryotes). Hsp and rRNA
sequences provide convincing evidence that organellar
genes are more closely related to their prokaryotic, than
their eukaryotic, counterparts.
The discovery of mitochondrion-like Hsp genes in earlydiverging, amitochondriate eukaryotes is only suggestive
evidence that mitochondria were present in the last
common ancestor of all extant eukaryotes. The origins of
the trichomonad hydrogenosome and mitochondrial Hsp
genes are presumably tied to one or more endosymbiotic
events. Some of those events, however, could have
occurred just after the separation of diplomonads from
other amitchondriate taxa; the branching patterns for the
deepest-diverging eukaryotic taxa are still unresolved.
Furthermore, we do not know the identity of the earliestdiverging eukaryotic lineages. That we know the phylogenetic position of trichomonads, diplomonads and
microsporidia is a historical accident, guided by the
1. Müller M: The hydrogenosome. J Gen Microbiol 1993,
139:2879–2889.
2. Patterson DJ, Sogin ML: Eukaryote origins and protistan diversity.
In The Origin and Evolution of Prokaryotic and Eukaryotic Cells.
Edited by Hartman H, Matsuno K. New Jersey: World Scientific
Publishing Company; 1992:13–46.
3. Leipe D, Gunderson JH, Nerad TA, Sogin ML: Small subunit
ribosomal RNA of Hexamita inflata and the quest for the first
branch in the eukaryotic tree. Mol Biochem Parasitol 1993,
59:41–48.
4. Knoll AH: The early evolution of eukaryotes: a geological
perspective. Science 1992, 256:622–627.
5. Horner DS, Hirt RP, Kilvington S, Lloyd D, Embley TM: Molecular
data suggest an early acquisition of the mitochondrion
endosymbiont. Proc R Soc London [Biol] 1996, 263:1053–1059.
6. Bui ETN, Bradley PJ, Johnson PJ: A common evolutionary origin for
mitochondria and hydrogenosomes. Proc Natl Acad Sci USA
1996, 93:9651–9656.
7. Germot A, Philippe H, Guyader HL: Presence of a mitochondrialtype 70-kDa heat shock protein in Trichomonas vaginalis
suggests a very early mitochondrial endosymbiosis in eukaryotes.
Proc Natl Acad Sci USA 1996. 93:14614–14617.
8. Roger AJ, Clark CG, Doolittle WF: A possible mitochondrial gene in
the early-branching amitochondriate protist Trichomonas
vaginalis. Proc Natl Acad Sci USA 1996. 93:14618–14622.
9. Müller M: Evolutionary origins of trichomonad hydrogenosomes.
Parasitol Today 1997, in press.
10. Clark CG, Roger AJ: Direct evidence for secondary loss of
mitochondria in Entamoeba histolytica. Proc Natl Acad Sci USA
1995, 92:6518–6521.
11. Müller M: The hydrogenosome. In The Eukaryotic Microbial Cell.
Society for General Microbiology Symposium 30. Edited by Gooday
GW, Lloyd D, Trinci APJ. Cambridge, UK: Cambridge University
Press; 1980:127–142.